videomae-base-finetuned-subset-0401

This model is a fine-tuned version of MCG-NJU/videomae-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6379
• Accuracy: 0.7824

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 2775

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.6048	0.02	56	1.6213	0.0829
1.5891	1.02	112	1.5230	0.2811
1.4797	2.02	168	1.6437	0.1982
1.3999	3.02	224	0.9263	0.7465
1.0917	4.02	280	1.2308	0.4931
1.238	5.02	336	0.9406	0.6590
1.1525	6.02	392	0.8809	0.7051
1.0806	7.02	448	1.0089	0.5945
0.8483	8.02	504	0.9700	0.5853
0.992	9.02	560	1.1880	0.4885
0.862	10.02	616	0.7174	0.7512
1.0694	11.02	672	0.8598	0.7143
0.8885	12.02	728	0.8290	0.7097
0.8965	13.02	784	0.8304	0.7143
0.7371	14.02	840	0.7009	0.7696
0.6872	15.02	896	0.6768	0.7926
0.6022	16.02	952	0.7513	0.7373
0.9308	17.02	1008	0.8055	0.7097
0.4456	18.02	1064	0.7876	0.6728
0.6802	19.02	1120	0.7224	0.7235
0.7154	20.02	1176	0.7434	0.7051
0.503	21.02	1232	0.8346	0.6959
0.7203	22.02	1288	0.9694	0.5991
0.6799	23.02	1344	0.6474	0.7696
0.5802	24.02	1400	0.9573	0.6359
0.7047	25.02	1456	0.9120	0.6959
0.6701	26.02	1512	1.1690	0.5853
0.5514	27.02	1568	0.9174	0.6866
0.538	28.02	1624	0.8543	0.6866
0.7226	29.02	1680	0.7774	0.7465
0.4459	30.02	1736	0.9135	0.6359
0.3905	31.02	1792	0.8586	0.6728
0.7071	32.02	1848	0.7919	0.7327
0.4983	33.02	1904	0.7507	0.7512
0.5654	34.02	1960	0.7679	0.7143
0.5569	35.02	2016	0.8438	0.7097
0.3998	36.02	2072	0.8691	0.7189
0.5341	37.02	2128	0.8056	0.7604
0.4024	38.02	2184	0.7071	0.7880
0.5011	39.02	2240	0.8827	0.7005
0.5857	40.02	2296	0.8525	0.7097
0.5619	41.02	2352	0.8228	0.7512
0.6052	42.02	2408	0.8320	0.7373
0.5124	43.02	2464	0.8776	0.7419
0.3323	44.02	2520	0.8515	0.7465
0.5684	45.02	2576	0.9309	0.7097
0.4406	46.02	2632	0.8826	0.7465
0.6164	47.02	2688	0.8994	0.6959
0.4549	48.02	2744	0.8700	0.7189
0.3453	49.01	2775	0.8822	0.7189

Framework versions

• Transformers 4.36.2
• Pytorch 1.13.1
• Datasets 2.16.1
• Tokenizers 0.15.0